III. MATERIALES Y MÉTODOS
3.8. Análisis de los resultados
3.8.2 Confiabilidad del instrumento desempeño docente para la variable
Two issues are subsumed under this section. The first relates to the convergent and predictive validity of the assessment battery, that is the evidence that performance on the battery relates to, and is predictive of, academic achievement. The second issue concerns the extent to which performance on the battery can provide guidance for appropriate remediation.
Predictive validitv
Das and Mesnick investigated the link between processing style and school achievement, in a group of 79 fifth grade children (Das & Mesnick 1989). Achievement was measured using the routine assessments made of literacy and
mathematics by the particular educational district in which the study took place. As per their hypothesis, reading achievement was predicted by high scores in both simultaneous and sequential processing, with different aspects of reading related to the different processing styles. Mathematical achievement, too, was predicted by both simultaneous and sequential processing, with a high simultaneous score being the more closely associated (Das & Mesnick 1989).
The K-ABC has also been successfully applied to the prediction of academic achievement amongst children coming from different cultural backgrounds (Worthington & Bening 1988). Children included in the study were between 8 to 12 years old, and had all been referred for special educational services for either mathematical or literacy support (N = 99). They were equally divided amongst three cultural groups: African-American; Hispanic; and White. Eight simultaneous -sequential sub-tests were administered, and a Mental Processing Composite score (MPC) was calculated for each child. From this a mean MPC score was derived for each of the three sub-groups. Regression analyses were carried out, with the MPC score being used as the predictor variable for academic achievement (measured by the reading, mathematics, and written language sections of the Woodcock-Johnson Psycho-Educational Achievement Battery). The correlations between K-ABC scores and the achievement measures ranged from R = .2 to .44. Although significant, these were lower than those reported in the K-ABC Interpretive Manual, and this was attributed to having a sample population with learning difficulties. No significant differences were found in the pattern of correlations between any of the three groups (Worthington & Bening 1988).
Appropriate Remediation: Aotitude-Treatment-lnteraction
Given the association found between processing strengths and academic
achievement {Das & Mesnick 1989), it might be assumed that different academic
difficulties w\W also relate to different processing skills (Bain 1993). In this respect, despite stressing the importance of forging a link between processing style and teaching style (Kaufman & Kaufman 1983), the K-ABC does not actually allow an analysis of the processes by which a child solves the problem. Rather it makes
assumptions about the processes that should be applied, and then attempts to quantify them. The inherent assumption is that poor performance is indicative of the application of an inefficient strategy, and thus by implication a preference for the inappropriate strategy (Kaufman & Kaufman 1983, Sternberg 1984a, Das 1984, Das et al 1990). The corollary to this is that the greatest improvement in academic performance should be found to follow the matching of teaching style to the child's processing strength (Kaufman & Kaufman 1983). The link is described as the ATI, Aptitude-Treatment-lnteraction paradigm (Cronbach cited in Kaufman & Kaufman 1983 p.242). The approach recommended by Kaufman combines task analysis, which identifies what to teach, with the most appropriate way to teach it, through the identification of the child's aptitude (processing strength).
A number of studies have been carried out to investigate the ATI link, some supportive and some sceptical (Gunnison 1982, Ayres & Cooley 1986, Ayres et al 1988, Kempa et al 1988, Bain 1993, Good et al 1993). AH these studies suffer from methodological weaknesses, foremost of which is their small sample sizes (N = 6 - 44). However they raise the following issues concerning the concept of ATI and the K-ABC as a tool by which to apply it. The data suggests that :
1. problems associated with ATI procedures exist, and that :
"These problems include the accurate assessment of specific aptitudes that are linked directly to specific instructional procedures. Without valid evidence of the connections between the measurement of aptitudes and the provision of teaching procedures, the relationship between a specific aptitude we attempt to measure and a like-labelled teaching procedure we devise may well be tenuous. " (Ayres et al 1988 p.123)
2. the K-ABC is a valid instrument for :
"investigating the nature of processing styles in the learning disabled population". (Kempa et al 1988 p.242)
b u t , little support was found for :
"making more specific recommendations about appropriate remedial programs". (Kempa et al 1988 p.242).
3. because of
"the possibility of existing sub-population processing style differences in terms of culture and social class. " (Bain 1993 p.243)
interpretation of individual processing strengths need to be made in the context of group norms, and the definition of what constitutes a homogeneous group also requires careful analysis (Bain 1993).
Thus, whilst there is evidence that performance on the K-ABC relates directly to academic achievement (Worthington et al 1988; Das et al 1989), still much remains to be understood about the relationship between the measure of specific aptitude and the design of teaching procedures (Ayres et al 1988; Bain 1993). if, and how, knowledge about processing strength can be exploited to provide guidelines for remedial programmes requires further investigation (Good et al 1993).
Goals 4,5: Dealing With Novelty
In avoiding the measurement of acquired knowledge in the mental processing scales, and by including novel tasks, the K-ABC has addressed an important aspect of intellectual functioning, namely the ability to deal with novel items (Sternberg 1984b). This in turn has received the commendation of an otherwise severe critic (Sternberg 1984a). Also commended was the test administration procedure, where the inclusion of teaching items was aimed at increasing the child's familiarity with the testing procedure (Sternberg 1984a). These aspects of the K-ABC bring this otherwise "static" test in closer alignment with examples of "dynamic" assessment such as Feuerstein's Learning Potential Assessment Device (LPAD) (Kaufman & Kaufman 1983, Feuerstein et al 1987, Lidz & Thomas 1987, and a more detailed discussion in Chapter 9).
The K-ABC essentially falls into the category of a static test, and has the advantage of providing a standardised format in order to assure compatibility and reliability. The dynamic aspects incorporated into its design include the provision, at the beginning of each sub-test, to give demonstration and instruction for those
children who do not successfully complete the first items. Both the K-ABC and the LPAD therefore incorporate teaching children "who have not grasped certain concepts" (Kaufman & Kaufman 1983), although, as Kaufman points out, "the reasons for teaching are different " The teaching in the LPAD allows for a detailed analysis of how the child approaches the solution to the problem, and aims at beginning the process of structural change. Feuerstein refers to the form of teaching adopted in the K-ABC, however, as placing an
" emphasis on affecting peripheral components of the individual's functioning (by) offering a restricted experience with tasks similar to those that are the criterion for success in the test itself" (Feuerstein et al 1987 p.44).
Both the LPAD and the K-ABC also place an emphasis upon the inclusion of novel items. However the LPAD was designed for use with older children, and the novel items it includes tend to be more abstract than those included in the K- ABC (Kaufman & Kaufman 1983). Consequently the content of the K-ABC is seen as more appropriate for younger children (Lidz & Thomas 1987).
These aspects of the format of the K-ABC makes it particularly attractive for use with populations who have little or no previous experience of one-one testing formats (Cline 1992; Gamlin & Luther 1992, Giordani et al 1996).
Goal 6: Sensitivity To Different Groups Of Children